文件列表:

LICENSE (1077, 2023-09-26)
assets/ (0, 2023-09-26)
assets/robotic werewol 0.jpg (241040, 2023-09-26)
core/ (0, 2023-09-26)
core/__init__.py (0, 2023-09-26)
core/api.py (1508, 2023-09-26)
core/game.py (11124, 2023-09-26)
docs/ (0, 2023-09-26)
docs/Alternatives.md (151, 2023-09-26)
docs/Samples/ (0, 2023-09-26)
docs/Samples/1.md (29725, 2023-09-26)
docs/image-1.png (89076, 2023-09-26)
docs/index.md (699, 2023-09-26)
log-to-readable-md.py (2857, 2023-09-26)
log.json (33108, 2023-09-26)
mkdocs.yml (128, 2023-09-26)
prompts/ (0, 2023-09-26)
prompts/__init__.py (0, 2023-09-26)
prompts/render_prompts.py (4970, 2023-09-26)
requirements.txt (22, 2023-09-26)
run.py (136, 2023-09-26)

# MafiaGPT a program that moderate a [Mafia(Werewolf)](https://en.wikipedia.org/wiki/Mafia_(party_game)) game with 7 players between GPT models. to see communication, team work and social/logical abilities of GPT model.(it is fun and educational) the roles are : - 1 Medical - 1 Seer - 3 Villagers - 2 Werewolves ## Get started install openai via pip : ```shell pip install openai ``` set your Openai API key as an enivronmental variable with key : OPENAI_API_KEY - you can configure your model(Default is 'gpt-3.5-turbo') and your rate_limit and token_limit in api.py - make DEBUG=True in api.py if you want to see all messages passed to GPT model. ```shell python run.py ``` it will save a log file in the bellow format in records/ folder. ## log.json format every event in the game is a dictionary. ```javascript { 'event':"", 'content':"<...>" } ``` events are : - roles : ```javascript 'content':{ 'player':"" 'role':"" } ``` - cycle : ```javascript 'content':"" ``` - speech : ```javascript 'content':{ 'player':"" 'context':"" } ``` - voted : ```javascript 'content':{ 'player':"" 'voted_to_player':"" 'reason':"" } ``` - vote_start : ```javascript 'content':{ 'player':"" 'voted_to_player':"" 'reason':"" } ``` - healed : ```javascript 'content':{ 'player':"" 'reason':"" } ``` - targeted : ```javascript 'content':{ 'player':"" 'reason':"" } ``` - killed : ```javascript 'content':{ 'player':"" } ``` - inquiried : ```javascript 'content':{ 'player':"" 'context':"" 'reason':"" } ``` - notetaking : ```javascript 'content':{ 'player':"" 'context':"" } ``` - end : ```javascript 'content':{ 'winner':"" } ``` ## Prompts - Game Introduction('/intro_prompt.txt') : rules and the role of the player - Game Reports('/report_prompt.txt) : - Speeches in last round - Game status : alive players in seperation of the teams. - Players status : alivness of players based on their number - Agent previous SPEACIAL actions - Agent previous VOTES - Notes that have taken by the agent - Commands('/game.py') : - Speak command - Vote command - targetting/healing/killing/inquiring command ## Tasks - [ ] optimizing prompts - [ ] a web interface to play natively a log-file as an animation - [ ] allowing a player/players to play with GPT (Probably tricky prompts will lead to game hacking.) - [ ] allowing more chat models or different GPT tempretures to compete

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：